Andrew James ROBBIE wrote:
> 
> I've tried out a couple of the recent kernels released as snapshots, but
> none of them them successfully boot. This is on a Mac IIvx, 20MB RAM,
> Apple 2-page monochrome, Apple standard ADB devices. 1.3.2 runs fine,
> so it isn't a major hassle, but I would like to try out -current before
> the 1.4 release...

well, i don't think there's a lot of hurry at the moment.  we certainly
haven't started the release cycle yet.
 
> This is the boot sequence from the 19980905 snapshot (using GENERIC and
> MRG kernels produces the same results) :
> 
>     - MacOS 7.1.2: Booter 1.11.3. Booting off the -current kernel sitting
>       on my 1.3.2 root partition. Disk is a quantum 250MB.
>     - The booter loads the kernel file and starts NetBSD without any
>       problems.
>     - NetBSD boot sequence comes up, messages as usual until
> 
>       "asc0 at obio0: Apple Sound Chip" (just after "adb: 100 dpi mouse at 3")
> 
>       at which point it freezes. Pressing the hardware interrupt button:
> 
>       PC is 0x13676e
> 
>       and doing a stack trace
> 
[snip]
>         _lev2intr(
>         _asc_intr_enable(
>         _ascattach(
[snip]

it's hanging while setting up the ASC's interrupt handler.  that's very
strange.  of course, what's even stranger is that you're actually getting
a level 2 interrupt at the same time....without looking at the code, i
can't remember what else is on the second VIA controller....perhaps the
SCSI....

> Does anyone know what could be going wrong?

you might be hitting some kind of funky race condition with your SCSI
drive, but i'm really not sure on that one.

> I'm happy to try out any
> special debug kernels if that would help.

can you compile you're own kernels?  if so, i might be able to come up
with a debug patch so that we can see what's going on...maybe we just need
to turn off interrupts while we set up the ASC interrupt handler (or are
they already off?).

> Also, I am wondering if it
> could be related to the problem reported by Hans-Christian Becker
> (Thu, 10 Sep 1998 15:31:59 +0200, "-current & GENERIC-84 problems")
> where it stops at "Configuring network interfaces: mc0". I don't have
> any ethernet cards connected, it just seems slightly similar.

nope, that was due to an interrupt handling problem with the AV Quadras.
it shouldn't affect II-series machines.  of course it's possible that
there is still an interrupt handling problem, but the AV one is now fixed.

later.

-- 
Colin Wood                                 cwood@ichips.intel.com
Component Design Engineer - PMD                 Intel Corporation
-----------------------------------------------------------------
I speak only on my own behalf, not for my employer.